Fix flaky Retry-After Header Test

[travis-minke-sap]

Fixes #5937
(hopefully - unable to reproduce locally - deducing timing mismatch root cause indirectly ; )

Proposed Changes

• 🐛 Refactor the RetryAfterValidationServer used in test for tracking request intervals.

Pre-review Checklist

• At least 80% unit test coverage
• E2E tests for any new behavior. (n/a changes are to unit test only)
• Docs PR for any user-facing impact (n/a changes are to unit test only)
• Spec PR for any new API feature (n/a changes are to unit test only)
• Conformance test for any change to the spec (n/a RetryAfter support is an alpha level experimental-feature and does not yet have Conformance tests.)

Dev Notes

The test/cases in question have never been seen to fail locally so the root cause is not 100% certain. The changes to the test are an attempt to address the timing concerns in a more robust manner.

The test case in question involves returning the Retry-After headers in the RFC850 date format. This formatting drops/truncates milliseconds and thus requires special handling in the test.

The prior implementation was rounding the request receipt times in an effort to work around this. This is risky as either request (current/previous) might round towards or away from each other making the interval checking difficult.

The new approach instead uses the actual request receipt times, and simply adjusts the acceptable time-window when dealing with these RFC850 date test cases. This should be more robust as it is not impacted by all the various rounding permutations.

[codecov]

Codecov Report

Merging #5943 (7a90a9b) into main (e73a69b) will not change coverage.
The diff coverage is n/a.

@@           Coverage Diff           @@
##             main    #5943   +/-   ##
=======================================
  Coverage   82.23%   82.23%           
=======================================
  Files         220      220           
  Lines        7572     7572           
=======================================
  Hits         6227     6227           
  Misses        911      911           
  Partials      434      434           

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e73a69b...7a90a9b. Read the comment docs.

[matzew]

/lgtm
/approve

[knative-prow-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: matzew, travis-minke-sap

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [matzew]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[pierDipi]

Thanks!